about summary refs log tree commit homepage
DateCommit message (Collapse)
2019-07-07nntp: improve error reporting for COMPRESS
Add some checks for errors at initialization, though there's not much that can be done with ENOMEM-type errors aside from dropping clients. We can also get rid of the scary FIXME for MemLevel=8. It was a stupid error on my part in the original per-client deflate stream implementation calling C::R::Z::{Inflate,Deflate} in array context and getting the extra dualvar error code as a string result, causing the {zin}/{zout} array refs to have extra array elements.
2019-07-06nntp: reduce memory overhead of zlib
Using Z_FULL_FLUSH at the right places in our event loop, it appears we can share a single zlib deflate context across ALL clients in a process. The zlib deflate context is the biggest factor in per-client memory use, so being able to share that across many clients results in a large memory savings. With 10K idle-but-did-something NNTP clients connected to a single process on a 64-bit system, TLS+DEFLATE used around 1.8 GB of RSS before this change. It now uses around 300 MB. TLS via IO::Socket::SSL alone uses <200MB in the same situation, so the actual memory reduction is over 10x. This makes compression less efficient and bandwidth increases around 45% in informal testing, but it's far better than no compression at all. It's likely around the same level of compression gzip gives on the HTTP side. Security implications with TLS? I don't know, but I don't really care, either... public-inbox-nntpd doesn't support authentication and it's up to the client to enable compression. It's not too different than Varnish caching gzipped responses on the HTTP side and having responses go to multiple HTTPS clients.
2019-07-06nntp: support COMPRESS DEFLATE per RFC 8054
This is only tested so far with my patches to Net::NNTP at: https://rt.cpan.org/Ticket/Display.html?id=129967 Memory use in C10K situations is disappointing, but that's the nature of compression. gzip compression over HTTPS does have the advantage of not keeping zlib streams open when clients are idle, at the cost of worse compression.
2019-07-06nntp: move LINE_MAX constant to the top
It'll be accessible from other places, and there was no real point in having it declared inside a sub.
2019-07-06nntp: use msg_more as a method
It's a tad slower, but we'll be able to subclass this to rely on zlib deflate buffering. This is advantageous for TLS clients since (AFAIK) IO::Socket::SSL/OpenSSL doesn't give us ways to use MSG_MORE or writev(2) like like GNUTLS does.
2019-07-06watch: allow multiple spam watch directories
Given most folks have multiple mail accounts, there's no reason we can't support multiple Maildirs.
2019-07-06watch: remove some indirectly-used imports
We can drop some unnecessary imports and now that we switched to InboxWritable.
2019-07-05viewdiff: do not anchor using diffstat comments
Diffstat summary comments were added to git last year and we need to filter them out to get anchors working properly. Reported-by: SZEDER Gábor <szeder.dev@gmail.com> https://public-inbox.org/meta/20190704231123.GF20404@szeder.dev/
2019-07-05t/nntpd*.t: require IO::Socket::SSL 2.007 for Net::NNTP tests
Net::NNTP won't attempt to use older versions of IO::Socket::SSL because 2.007 is the "first version with default CA on most platforms" according to comments in Net::NNTP. But then again we don't make remote requests when testing...
2019-07-04qspawn: retry sysread when parsing headers, too
We need to ensure the BIN_DETECT (8000 byte) check in ViewVCS can be handled properly when sending giant files. Otherwise, EPOLLET won't notify us, again, and responses can get stuck. While we're at it, bump up the read-size up to 4096 bytes so we make fewer trips to the kernel.
2019-06-30Merge remote-tracking branch 'origin/nntp'
* origin/nntp: nntp: add support for CAPABILITIES command nntp: remove DISABLED hash checks
2019-06-30nntp: add support for CAPABILITIES command
Some clients may rely on this for STARTTLS support.
2019-06-30nntp: remove DISABLED hash checks
Before I figured out the long_response API, I figured there'd be expensive, process-monopolizing commands which admins might want to disable. Nearly 4 years later, we've never needed it and running a server without commands such as OVER/XOVER is unimaginable.
2019-06-30t/httpd-unix.t: avoid race in between bind() and listen()
We need to be able to successfully connect() to the socket before attempting further tests. Merely testing for the existence of a socket isn't enough, since the server may've only done bind(), not listen().
2019-06-30daemon: warn on inheriting blocking listeners
For users relying on socket activation via service manager (e.g. systemd) and running multiple service instances (@1, @2), we need to ensure configuration of the socket is NonBlocking. Otherwise, service managers such as systemd may clear the O_NONBLOCK flag for a small window where accept/accept4 blocks: public-inbox-nntpd@1 |systemd |public-inbox-nntpd@2 --------------------------+----------------+-------------------- F_SETFL,O_NONBLOCK|O_RDWR | | (not running, yet) |F_SETFL, O_RDWR | |fork+exec @2... | accept(...) # blocks! | |(started by systemd) | |F_SETFL,O_NONBLOCK|O_RDWR | |accept(...) non-blocking It's a very small window where O_NONBLOCK can be cleared, but it exists, and I finally hit it after many years.
2019-06-30tests: common tcp_server and unix_server helpers
IO::Socket:*->new options are verbose and we can save a bunch of code by putting this into t/common.perl, since the related spawn_listener stuff is already there.
2019-06-30t/perf-nntpd.t: fix off-by-one if NEWNEWS_DATE is unset
20190431 isn't real, NNTP.pm failed to parse it when our test client sent it.
2019-06-30Merge remote-tracking branch 'origin/email-simple-mem' into master
* origin/email-simple-mem: nntp: reduce syscalls for ARTICLE and BODY mbox: split header and body processing mbox: use Email::Simple->new to do in-place modifications nntp: rework and simplify art_lookup response
2019-06-30examples/*@.service: sockets MUST be NonBlocking
For users running multiple (-nntpd@1, -nntpd@2) instances of either -httpd or -nntpd via systemd to implement zero-downtime restarts; it's possible for a listen socket to become blocking for a moment during an accept syscall and cause a daemons to get stuck in a blocking accept() during PublicInbox::Listener::event_step (event_read in previous versions). Since O_NONBLOCK is a file description flag, systemd clearing O_NONBLOCK momentarily (before PublicInbox::Listener::new re-enables it) creates a window for another instance of our daemon to get stuck in accept(). cf. systemd.service(5)
2019-06-30ds: fix return values of do_read and *_tls_step
We need to ensure all these subroutines return false on incomplete.
2019-06-30ds: rely on refcounting to close descriptors
Since we have EPOLL_CTL_DEL implemented for the poll(2) and kqueue backends, we can rely on Perl refcounting to gently close(2) the underlying file descriptors as references get dropped. This may be beneficial in the future if we want to drop a descriptor from the event loop without actually closing it.
2019-06-30t/nntpd*.t: skip TLS tests for old Net::NNTP
Perl prior to 5.22 did not bundle a Net::NNTP (or libnet) capable of handling TLS.
2019-06-29dskqxs: more closely match epoll semantics
EV_DISPATCH is actually a better match for EPOLLONESHOT semantics than EV_ONESHOT in that it doesn't require EV_ADD for every mod operation. Blindly using EV_ADD everywhere forces the FreeBSD kernel to do extra allocations up front, so it's best avoided.
2019-06-29http: use bigger, but shorter-lived buffers for pipes
Linux pipes default to 65536 bytes in size, and we want to read external processes as fast as possible now that we don't use Danga::Socket or buffer to heap. However, drop the buffer ASAP if we need to wait on anything; since idle buffers can be idle for eons. This lets other execution contexts can reuse that memory right away.
2019-06-29httpd/async: switch to buffering-as-fast-as-possible
With DS buffering to a temporary file nowadays, applying backpressure to git-http-backend(1) hurts overall memory usage of the system. Instead, try to get git-http-backend(1) to finish as quickly as possible and use edge-triggered notifications to reduce wakeups on our end.
2019-06-29parentpipe: make the ->close call more obvious
We can close directly in event_step without bad side effects, and then we also don't need to take a reason arg from worker_quit, since we weren't logging it anywhere.
2019-06-29parentpipe: document and use one-shot wakeups
The master process only dies once and we close ourselves right away. So it doesn't matter if it's level-triggered or edge-triggered, actually, but one-shot is most consistent with our use and keeps the kernel from doing extra work.
2019-06-29http: support HTTPS (kinda)
It's barely any effort at all to support HTTPS now that we have NNTPS support and can share all the code for writing daemons. However, we still depend on Varnish to avoid hug-of-death situations, so supporting reverse-proxying will be required.
2019-06-29ds: consolidate IO::Socket::SSL checks
We need to be careful about handling EAGAIN on write(2) failures deal with SSL_WANT_READ vs SSL_WANT_WRITE as appropriate.
2019-06-29ds: handle deferred DS->close after timers
Our hacks in EvCleanup::next_tick and EvCleanup::asap were due to the fact "closed" sockets were deferred and could not wake up the event loop, causing certain actions to be delayed until an event fired. Instead, ensure we don't sleep if there are pending sockets to close. We can then remove most of the EvCleanup stuff While we're at it, split out immediate timer handling into a separate array so we don't need to deal with time calculations for the event loop.
2019-06-29listener: use edge-triggered notifications
We don't need extra wakeups from the kernel when we know a listener is already active.
2019-06-29http: use requeue instead of watch_in1
Don't use epoll or kqueue to watch for anything unless we hit EAGAIN, since we don't know if a socket is SSL or not.
2019-06-29ds: move requeue logic over from NNTP
We'll be reusing requeue in other places to reduce trips to the kernel to retrieve "hot" descriptors.
2019-06-29ds: share lazy rbuf handling between HTTP and NNTP
Doing this for HTTP cuts the memory usage of 10K idle-after-one-request HTTP clients from 92 MB to 47 MB. The savings over the equivalent NNTP change in commit 6f173864f5acac89769a67739b8c377510711d49, ("nntp: lazily allocate and stash rbuf") seems down to the size of HTTP requests and the fact HTTP is a client-sends-first protocol where as NNTP is server-sends-first.
2019-06-29t/ds-leak: fix race
We need to ensure we run lsof on the sleep(1) process, and not the fork of ourselves before execve(2). This race applies when we're using the default pure-Perl spawn() implementation.
2019-06-27watchmaildir: show the current path on spamcheck failures
Knowing which message failed a spam check is tough when I have many Maildirs and don't have a search indexing tool setup for spam mail.
2019-06-27nntp: reduce syscalls for ARTICLE and BODY
Chances are we already have extra buffer space following the expensive LF => CRLF conversion that we can safely append an extra CRLF in those places without incurring a copy of the full string buffer. While we're at it, document where our pain points are in terms of memory usage, since tracking/controlling memory use isn't exactly obvious in high-level languages. Perhaps we should start storing messages in git as CRLF...
2019-06-27mbox: split header and body processing
When dealing with ~30MB messages, we can save another ~30MB by splitting the header and body processing and not appending the body string back to the header. We'll rely on buffering in gzip or kernel (via MSG_MORE) to prevent silly packet sizes.
2019-06-27mbox: use Email::Simple->new to do in-place modifications
Email::Simple->new will split the head from the body in-place, and we can avoid using Email::Simple::body. This saves us from holding an extra copy of the message in memory, and saves us around ~30MB when operating on ~30MB messages.
2019-06-27nntp: rework and simplify art_lookup response
We don't need some of the array elements returned from art_lookup, anymore (and haven't used them in years). We can also shorten the lifetime of the Email::Simple object by relying on the fact Email::Simple->new will modify it's arg if given a SCALARREF and allow us to avoid Email::Simple::body calls. Unfortunately, this doesn't seem to provide any noticeable improvement in memory usage when dealing with a 30+ MB test message, since our previous use of ->body_set('') was saving some memory, but forcing a LF-only body to be CRLF was making Perl allocate extra space for s///sg.
2019-06-26certs/create-certs: create certs in 'certs/' directory
If running in our top-level source tree, use the 'certs/' directory as the prefix so we can just invoke `./certs/create-certs.perl' instead of `(cd certs && ./create-certs.perl)'
2019-06-26ds: cleanup poll test and avoid clobbering imports
On Linux systems with epoll support, we don't want to be clobbering defined subs in the t/ds-poll.t test; so use OO ->method dispatch instead and require users to explicitly import subs via EXPORT_OK.
2019-06-26Merge remote-tracking branch 'origin/nntp-tls'
* origin/nntp-tls: (59 commits) ds: ->write must not clobber empty wbuf array Makefile: skip DSKQXS in global syntax check ds: reduce overhead of tempfile creation Revert "ci: require IO::KQueue on FreeBSD, for now" ds: reimplement IO::Poll support to look like epoll ds: split out IO::KQueue-specific code daemon: use FreeBSD accept filters on non-NNTP daemon: set TCP_DEFER_ACCEPT on everything but NNTP nntp: send greeting immediately for plain sockets ci: require IO::KQueue on FreeBSD, for now nntp: lazily allocate and stash rbuf ds: flush_write runs ->write callbacks even if closed nntp: simplify long response logic and fix nesting ds: always use EV_ADD with EV_SET nntp: reduce allocations for greeting ds: allow ->write callbacks to syswrite directly daemon: use SSL_MODE_RELEASE_BUFFERS t/nntpd-tls: slow client connection test nntp: call SSL_shutdown in normal cases ds|nntp: use CORE::close on socket ...
2019-06-25searchview: avoid displaying full paths on errors
Displaying full path names of installed modules could expose unnecessary information about user home directory names or other potentially sensitive information. However, displaying a module name could still be useful for diagnosing problems, so map full paths to the relevant part of the path name which is relevant to the package name. Reported-by: Ali Alnubani <alialnu@mellanox.com> https://public-inbox.org/meta/20190611193815.c4uovtlp574bid6x@dcvr/
2019-06-24msgmap: mid_insert: use plain "INSERT" to detect duplicates
"INSERT OR IGNORE" still bumps the auto-increment counter in SQLite, which causes gaps to appear in NNTP article numbering. This bug appeared in v2 repos where V2Writable may call ->add repeatedly on the same message. This bug is apparent with public-inbox-watch and work-in-progress IMAP watchers which may rescan and (attempt to) reinsert the same message on mailbox changes. Most uses of public-inbox-mda were not affected, unless the same message is actually delivered multiple times to the mda. v1 is not affected, either, since deduplication is only based on Message-ID and msgmap never sees the duplicate. Reported-by: "Eric W. Biederman" <ebiederm@xmission.com>
2019-06-24ds: ->write must not clobber empty wbuf array
We need to account for ->write(CODE) calls doing ->write(SCALARREF), otherwise flush_write may see the wrong ->{wbuf} field.
2019-06-24Makefile: skip DSKQXS in global syntax check
IO::KQueue isn't easily installable on Linux systems.
2019-06-24ds: reduce overhead of tempfile creation
We end up buffering giant things to the FS sometimes, and open() is not a cheap syscall; so being forced to do it twice to get a file description with O_APPEND is gross when we can just use O_EXCL ourselves and loop on EEXIST.
2019-06-24Revert "ci: require IO::KQueue on FreeBSD, for now"
Now that we support IO::Poll once again, we can remove the IO::KQueue requirement.
2019-06-24ds: reimplement IO::Poll support to look like epoll
At least the subset of epoll we use. EPOLLET might be difficult to emulate if we end up using it.